Data-driven methods to discover stable linear models of the helicity injectors on HIT-SIU
Zachary Daniel, Alan Kaptanoglu, Chris Hansen, Kyle Morgan, Steven L. Brunton, J. Nathan Kutz · 2025
Read Data-driven methods to discover stable linear models of the helicity injectors on HIT-SIU by Zachary Daniel, Alan Kaptanoglu, Chris Hansen, Kyle Morgan, Steven L. Brunton, J. Nathan Kutz (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-05405.txt.
Description
Zachary Daniel and colleagues use data-driven methods to build simple, stable circuit models for a plasma physics experiment. The device forms a ball of hot plasma and drives it with injector circuits whose behavior is tightly tied to plasma performance. Linear models are favored for control because their behavior is predictable, yet they miss nonlinear effects and shifts in settings. Writing in twenty twenty-five, the team applies a bagged, optimized version of dynamic mode decomposition, a method that learns a compact model straight from data while enforcing stability.
They train and test the method on analytic models of the circuits, with and without plasma, and then on real experimental shots. In each setting the learned models stayed stable and matched the system well. The authors present the approach as a practical tool for control and uncertainty checks in plasma experiments.
Catalog record
| CREATOR | Zachary Daniel, Alan Kaptanoglu, Chris Hansen, Kyle Morgan, Steven L. Brunton, J. Nathan Kutz |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 59 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-05405